I have a few clusters with 150-200+ nodes (EC2 instances) running ES 1.3.2. Looking for a 'faster' way to upgrade these clusters to ES 1.5.2. Following recommendations for rolling upgrades may take me a few days. Can not take downtime.
Any recommendations?
You will possibly get faster results by spinning up one or more extra nodes with the new version. You effectively pop out the old nodes as data is migrated to the new ones, then upgrade and re-add the "old" nodes back into the cluster.
Other than that, there is no substitute for rolling restarts of which I am familiar.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.